Article acceptance date: 10 February 2018

This R-Markdown document provides step-by-step instructions for executing the analysis and producing figures presented in the paper Quantifying soil moisture impacts on light use efficiency across biomes by Stocker et al. (2018). The starting point of the present collection of scripts is the data file ./data/modobs_fluxnet2015_s11_s12_s13_with_SWC_v3.Rdata which contains all the data used for the neural network (NN) training and analysis. Further information about data processing and creating the data file is given below (section ‘Data processing’). Using RMarkdown and open access code available through github (https://github.com/stineb/nn_fluxnet2015) this is supposed to allow for full reproducibility of published results - from publicly accessible data files to published figures.

IP information

All code in this repository is free software. It may be redistributed and/or modified under the terms of the GNU General Public License as published by the Free Software Foundation, version 3. Present code is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License (file ./LICENSE) for more details.

Copyright (C) 2017 Benjamin D. Stocker


We quantify the fractional reduction in light use efficiency due to soil moisture, separated from VPD and greenness effects as the ratio of actual versus potential LUE: \[ \mathrm{fLUE} = \mathrm{LUE}_{\mathrm{act}}\; /\; \mathrm{LUE}_{\mathrm{pot}} \] “Potential”" light use efficiency (LUE\(_{\mathrm{pot}}\)) is predicted using artificial neural networks (NN, see below), trained on the empirical relationship between observed LUE (LUE\(_{\mathrm{obs}}\)) and its predictors temperature, VPD, and PAR during days where soil moisture is not limiting (“moist days”). All NN training is done for each site specifically. “Actual” LUE (LUE\(_{\mathrm{act}}\)) is derived from NNs using all data and, in contrast to the NN for LUE\(_{\mathrm{pot}}\), with soil moisture as an additional predictor (see Fig. S1).